OpenAI to release Astra, its first model hitting 'critical' cyber capability threshold
nordicinst · x · 2026-09-02
WIRED reports OpenAI's forthcoming Astra is the company's first model to reach the "critical" cybersecurity threshold in its preparedness framework—meaning it can independently find and exploit previously unknown vulnerabilities in real-world software.
Per protocol, OpenAI paused some Astra-related training workloads for several weeks, resuming only after adding extra safety and security controls. Executives say they are now confident Astra can be released broadly and safely.
Astra will get a public release "soon," but its advanced cyber capabilities will be limited at launch to select partners in the Daybreak Blue early-access program, giving defenders time to shore up systems. The move comes as Silicon Valley grapples with advanced cyber capabilities of frontier models, with European AI players bracing for cyber-savvy rivals.
Related event: OpenAI Previews Astra as gpt-6-astra Traces Multiply Ahead of Launch(85 posts)→
More from Models
- Models can detect read-only probes with just last-two-layer SFT — Sauers_ · 2026-10-18
- Paying Google AI subscriber blocked: custom-domain Workspace accounts can't access AI Studio — ChrisRackauckas · 2026-10-18
- Running Ling 3.1 Flash (560B MoE, 25B Active) Across 4 DGX Sparks via llama.cpp RPC — BerryBlush_Rogers691 · 2026-10-18
- Users press OpenAI on long-delayed Adult Mode a year after it was promised — ___Patrice___ · 2026-10-18
- Polymarket opens betting market on who releases the next Gemini Argon model — Polymarket · 2026-10-18
- Grok matter-of-factly lists which groups user was racist against, exempting Africans and Latin Americans — teortaxesTex · 2026-10-18